AI-generated editorial illustration for the Deepfake nudes: first 48 hours guide — no real person depicted
AI-generated image (EU AI Act Art. 50 label). Editorial illustration — no real person depicted. See our AI content policy.

Deepfake Nudes of You: What to Do in the First 48 Hours

An hour-by-hour plan for adults who find an AI-generated intimate image of themselves online, based on the removal duties that platforms have carried since May 19, 2026.

Safety & LawUpdated: September 3, 2026Reading time: 6 minClothoff AI Editorial Team

A deepfake nude is an AI-altered intimate image of a real person, and since May 19, 2026 US platforms must remove it within 48 hours of a valid request under the TAKE IT DOWN Act. The plan below covers evidence, removal, hashing and reporting in order. It is written for adults; see our Responsible AI policy for consent rules.

What should happen in the first hour?

Do not engage with whoever posted the image, and do not pay anyone who threatens to spread it; extortion attempts escalate when they get a response. Instead, capture evidence before anything disappears: full-page screenshots showing the URL and date, the profile or channel name, and any message that reveals intent. Save the original photo the forgery was built from if you can identify it, because it proves the depiction is altered.

Then decide who else needs to know. If the person depicted is under 18, stop here and go straight to NCMEC’s CyberTipline; different rules and faster channels apply. For adults, the next steps are removal, hashing and reporting, roughly in that order. Our deepfake explainer describes how these images are produced, which helps when you write the notice.

How do you get the image removed?

Every covered platform, from social networks to image hosts, must offer a plain-language removal process and act within 48 hours of a valid notice. The FTC lists what a notice needs: your identity or an authorized agent, the location of the image, a statement that it is intimate and non-consensual, and a contact address.

HourActionWhere
0–1Screenshots with URL and timestamp; save linksYour device, offline copy
1–2Submit the platform’s removal request; note the confirmation numberPlatform form or Report NCII
2–4Create hashes so copies are blocked before they spreadStopNCII
4–24If the image came from a tool with a known removal route, file there tooProvider abuse mailbox, see reviews
24–48Escalate: FTC complaint if the platform ignores you; police report if threatenedTakeItDown.ftc.gov

TIDA requires covered platforms that receive a valid request to remove the content and known identical copies within 48 hours.

— FTC, Complying With the Take It Down Act, accessed Sep 3, 2026

Keep every confirmation number. If the platform misses the 48-hour window, that record is what the FTC complaint form asks for. Our NCII entry explains how the notice-and-removal duty differs from a copyright takedown.

Which tools have a working removal channel?

When the source tool is known, a direct complaint can stop regeneration and delete stored inputs. Among the six services reviewed at T0, only Ainudez publishes a TAKE IT DOWN complaints policy (dated May 6, 2026) with a 48-hour commitment. Pornworks offers a content-removal form with no stated deadline; Undresswith AI routes deletion through a support email; Deep Undress and UndressHer rely on in-account deletion and a contact form. Wavespeed states that generated media expire within 7 days. Full routes are collected in how to get an image removed from undress apps.

Who else should you report it to?

Three reporting lines run in parallel and none cancels another. The FTC handles platforms that fail their removal duty. Local police handle threats, extortion and, in states with criminal deepfake statutes, the creation or distribution itself; our state deepfake laws entry lists which statutes apply where you live. If the image appears in search results, Google’s removal request for explicit personal imagery is a separate route.

For hash-based prevention, StopNCII creates a fingerprint on your device and shares only the hash with participating platforms; the image never leaves your phone. For anyone under 18, NCMEC’s Take It Down does the same and the CyberTipline takes the report. Our StopNCII entry walks through the hashing steps.

Criminal complaints rest on the TAKE IT DOWN Act for publication and threats, with up to 2 years for adult victims, plus state statutes for creation where they exist. Civil claims are state-based for now: California and Illinois allow lawsuits, and one 2026 statute permits up to three times actual damages. The federal DEFIANCE Act, which would add a nationwide civil remedy, passed the Senate on January 13, 2026 but is still pending in the House as of September 3, 2026. Our legality guide has the statute table.

Sources and verification

Primary sources checked for this article. Service facts come from the linked editorial reviews and their dated test notes.

This article is editorial research, not legal advice. Laws change; statutes and agency pages were checked on September 3, 2026. For a specific situation, consult a licensed attorney in your state.

Consent and reporting

Frequently asked questions

Should I contact the person who posted the image?

No. Responding confirms the image reached you and often triggers demands for money. Document the account instead, block it after screenshots are saved, and route everything through the platform notice, the FTC form and, if threats are involved, the police. Extortion is a separate crime regardless of whether the image is real or forged.

What if the platform does not remove the image in 48 hours?

Keep the confirmation number and timestamps, then file a complaint at the FTC’s TAKE IT DOWN portal, which opened on May 19, 2026. The FTC treats a missed deadline as a violation with civil penalties up to $53,088 each. Re-submit the platform notice as well, because copies may have been posted under new URLs.

Does StopNCII work for AI-generated images?

Yes. StopNCII creates a hash of the image on your own device and shares only that fingerprint with participating platforms, which then block matching uploads. It works for altered and synthetic images of adults as long as you have a copy. It does not remove images already published; the platform notice does that.

Can I find out which app was used to make the image?

Sometimes. Watermarks, output dimensions and typical artifacts point to particular tools, and our reviews document each tool’s output signature. If you identify the service, file a complaint with its abuse channel so stored inputs are deleted. Do not create an account or pay the service to investigate; screenshots of its public pages are enough.

Is it still a crime if the deepfake is obviously fake?

Publication without consent is the offense; the statute covers any intimate depiction of an identifiable person that was created or altered with software. Quality does not matter, and a caption calling it fake does not cure it. Threats to publish are punishable even when nothing is posted, with up to 18 months for adult victims.

What if I am under 18 or the image shows a minor?

Stop and use NCMEC’s channels: Take It Down for hashing and the CyberTipline for the report. Any sexualized image of a minor is a federal crime whether or not it was generated by AI, and platforms treat these reports with priority. A trusted adult or school counselor can file on your behalf.